Prodrug gene therapy is a therapeutic strategy in which tumor cells are transfected with a 'suicide' gene that encodes a metabolic enzyme which is capable of converting a nontoxic prodrug into a potent cytotoxin. Such a method allows selective eradication of tumor cells while sparing normal tissue from significant cell killing. The effectiveness of this strategy is dependent on a bystander effect in which untransfected tumor cells are killed through active or passive transport of the cytotoxic enzyme product. Several enzyme/prodrug combinations are under active investigation, demonstrating effectiveness in both tissue culture and animal models. However, the combination of low transfection efficiencies and poor turnover of prodrug substrates limit the efficiency of cell killing in the tumor. In order to improve such therapies, enzyme variants must be selected and engineered for enhanced turnover of the prodrug substrate. In this proposal, a collaboration of two laboratories propose to optimize the nucleoside salvage enzymes cytosine deaminase and deoxycytidine kinase for prodrug suicide gene therapy, using a combination of structural biology and directed evolution screens, and to test the efficacy of enzyme variants in cell line and animal models. Cytosine deaminase (CD) catalyzes the deamination of cytosine to uracil and ammonia. Cytosine deaminase is found in bacteria and fungi but is not present in mammalian cells. Cells expressing CD are sensitive to the nucleoside analog, 5-fluorocytosine. Due to the enzymatic conversion of 5FC to 5-fluorouracil (5FU). This compound and its deoxyribonucleoside, fluorodeoxyuridine (FUdR), are potent inhibitors of DNA synthesis and RNA function and are widely used in cancer treatment. In contrast, deoxycytidine kinase (dCK) generates cytidine-monophosphate from cytidine nucleoside, and also activates the antineoplastic agents gemcitabine and cytarabine. The specific aims for this project are: (1) Determine the structure of bacterial CD, yeast CD and human dCK. (2) Perform structure-based mutagenesis and genetic screens to isolate enzyme variants with enhanced binding and turnover of prodrug substrates. This aim will exploit crystallographic information both to direct the mutagenesis of specific regions of the enzyme, and to directly visualize the structural basis of enhanced prodrug. activation by selected enzyme variants (3) Test enzyme variants for tumor cell killing in established tumor model systems, using cell lines and animal models. [unreadable] [unreadable]